archiving and accessing email from failed companies · background: user-driven inquiry ! how will...
TRANSCRIPT
Archiving and Accessing Email from Failed Companies David Kirsch Douglas Oard Jyothi Vinjumur University of Maryland http://ediscovery.umiacs.umd.edu/ Library of Congress Archiving Email Symposium June 2, 2015
Outline n Introduction & Background
n Collection Development
n Sample Collections n Ongoing & Next Steps
Background: User-driven inquiry
n How will we learn about what really goes on inside modern firms?
n Surviving firms unlikely to grant researchers access to large, un-vetted email collections
n Contra the Whig view that “history is written by the winners,” records of failures may be all that survives
n Failed firms are important, understudied historical agents
Source: Kirsch, “Record of Business at Risk” (2009)
Partners n Intermediaries
n Sherwood Partners (Workout specialist) n FTI Consulting (Bankruptcy Trustee) n Gallivan, Gallivan & O’Melia (Custodian)
n Sponsors n Sloan Foundation (BPA) n LoC / NDIIPP (DABODCE) n NSF (eDiscovery)
n Scholars n RRCHNM at GMU n UMD i-School / UMIACS n Advisory Council
n Repositories n Hagley Museum & Library (NDP) n Linguistic Data Consortium at UPenn n LC n Others?
Sample Collections: AvocadoIT
n Conducted risk assessment n Balance technical & institutional
mechanisms to mitigate risks
n Technical n Limited redaction (PII & images)
n Anonymization (AvocadoIT)
n Institutional n Limited access via LDC @ UPenn
n All users and institutions must sign license agreements
DOCUMENT TYPE COUNT PERCENT OFF TOTAL
Word Processor [.DOC, .WPF] 3195672 57%
Image Format [.TIF, .GIF] 831289 15%
System Configura&on Files [.DLL, .INI, …]
215274 5%
Email Message [.EML] 178503 3%
Spreadsheet [.XLS] 168206 3%
Text Format [.TXT, .RTF] 160438 3%
Database Files [.SQL,.MDB, …] 65709 1%
Error Log Files [.BAD] 66380 1%
Web Page [.HTM] 22343 < 1%
Portable Document [.PDF] 22287 < 1%
Data file [.DAT] 21614 < 1%
Miscellaneous [.LNK,.PST,.ZIP…] -‐ Approx. 10 -‐11%
SAMPLE COLLECTIONS: BROBECK [ DOCUMENT CATOGORIES ]
DOCUMENT TYPE COUNT PERCENT OFF TOTAL
Messages [No Extension] 935234 45%
Text Document [.TXT, .RTF] 859769 42%
Word Processor [.DOC, .WPF] 191725 9%
Portable Document [.PDF] 35345 2%
Spreadsheet [.XLS] 28634 1%
Image Format [.JPG, .GIF, .TIF, …] 12580 < 1%
Web Page [.HTM] 3895 << 1%
Power Point Presenta&on [.PPT] 3690 << 1%
Electronic Business Card [.VCF] 3590 << 1%
Link File iManage Format [.NRL] 2314 << 1%
Compressed file [.ZIP] 1252 << 1%
SAMPLE COLLECTIONS: BROBECK DESCRIPTIVE STATISTICS OF TOP 100 EXTRACTED .PST FILES [ DOCUMENT CATOGORIES ]
Ongoing & Next Steps n Ongoing
n Talking to LDC & Sherwood Partners about allowing commercial, as well as non-commercial, access to released collection
n Talking to Sherwood Partners & Hagley Library (NDP partner) to identify additional email collections
n Next n Talking to LoC to develop rules and
procedures for Brobeck collection
n Workshop to identify 3rd parties that can “pool” email collections?
Venture Workouts n Question: What happens to venture
backed firms that do not have successful exits? n Answer: Assignment for the Benefit of
Creditors or “Workout”—may be thought of as a private bankruptcy.
n Board of Directors replaces management with senior executive from specialized consulting firm that oversees liquidation. n Outsource non-core VC activity
n Specialized team increases value recovered
n Allows investors to maintain control (v. bankruptcy)
Sample – Sherwood Partners
n One of two Silicon Valley firms that service venture capital market. n Market share not known at present
n Handled several hundred workouts since late-1999 n Mostly IT though some biotech and medical
device companies.
n Seven year statute of limitations for records retention expiring and new failures arriving… n Sherwood needs to get rid of some of
7,000+ boxes